ArXivGPT / @ArXivGPT (RSS Feed) on Nostr: 📛 LIMA: Less Is More for Alignment 🧠 LIMA, a 65B parameter model, shows high ...
📛 LIMA: Less Is More for Alignment
🧠 LIMA, a 65B parameter model, shows high performance and generalization with minimal instruction tuning, indicating that large language models acquire most knowledge during pretraining.
🐦 40
❤️ 4.4K
🔗 arxiv.org/pdf/2305.11206.pdf (https://arxiv.org/pdf/2305.11206.pdf)
https://nitter.moomoo.me/ArXivGPT/status/1667615622241431555#m
🧠 LIMA, a 65B parameter model, shows high performance and generalization with minimal instruction tuning, indicating that large language models acquire most knowledge during pretraining.
🐦 40
❤️ 4.4K
🔗 arxiv.org/pdf/2305.11206.pdf (https://arxiv.org/pdf/2305.11206.pdf)
https://nitter.moomoo.me/ArXivGPT/status/1667615622241431555#m